Videoconference services provide two-way media data conferencing between multiple conference participants that may be associated with different networks, endpoints, or remote locations. Video processing is intensive and requires expensive dedicated hardware or significant quantity of commodity servers to scale to large numbers of conferences or participants. One videoconferencing technique is mesh-based videoconferencing wherein endpoints exchange streams of media data transmissions. Mesh-based videoconferencing requires significant network capabilities at participant endpoints, especially as the number of participants increases. Another videoconferencing technique uses a multi-connection unit (MCU) that reduces local strain on participant endpoints compared to mesh-based videoconferencing by handling the bulk of the data processing, transmission, and reception. MCU-based videoconferencing, however, merely reallocates the resource strain to the MCU service. The MCU is a central resource that limits the number of conferences or participants that can be held or joined concurrently to the capacity of the deployed MCU. As the number of participants and MCU videoconferences increase, the strain may affect the quality of the videoconference environment provided by the MCU service.